Oncofusion

start / graph / absence-q136-the-index-fusion-line-in-this-deposit-carries-h3k79me2-only-with-no-menin-chip-and-no-menin-inhibitor-arm

Analysisestablished confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.Q136-menin-vs-dot1l-inhibitors

CORRECTS A DISPATCH PREMISE: GSE149183's menin-versus-DOT1L-inhibitor head-to-head exists in MOLM13 (KMT2A::MLLT3) ONLY. RS4;11 - the index fusion - is in the series but carries H3K79me2 alone, with no VTP-50469 arm, no MEN1 ChIP and no DOT1L ChIP, and the two arms are on different genome builds

also called GSE149183 RS4;11 gap · the index-fusion arm that is not there

This dataset was expected to contain the key comparison in the exact leukaemia subtype the lab studies. It does not - that cell line was only used for a different measurement, so the main result has to be read as coming from a related but different leukaemia.

The same thing, for a scientist

**The dispatch brief for this beat stated that GSE149183 contains MEN1 and DOT1L ChIP under DMSO, VTP-50469, EPZ5676, compound 10 and compound 11 'in RS4;11 and MOLM13'. It does not.** All 31 sample records were fetched and enumerated [E1]: | samples | line | fusion | assays | drug arms | |---|---|---|---|---| | GSM4491616-GSM4491625 | MOLM13 | KMT2A::MLLT3 | **DOT1L ChIP, MEN1 ChIP** | DMSO, VTP-50469, EPZ5676, compound 10, compound 11 | | GSM4491626-GSM4491634 | MOLM13 | KMT2A::MLLT3 | H3K79me2 | DMSO, EPZ5676, compound 10 (d3/d6/d9) | | GSM4491635-GSM4491643 | **RS4;11** | **KMT2A::AFF1** | **H3K79me2 only** | DMSO, EPZ5676, compound 10 (d3/d6/d9) - **no VTP-50469** | | GSM4491615, GSM4491644, GSM4491645 | MOLM13 x2, RS4;11 | - | input | - | The series' own `overall_design` says it outright: *'Menin and DOT1L occupancy after treatment with DOT1L inhibitors and a Menin-inhibitor was assessed 4d after treatment with the drugs **in MOLM13 cells**'* [E1]. **Three consequences, in descending order of how much they cost.** **1. The head-to-head is partner-imported.** The strongest chromatin result this beat produced - menin cleared from every canonical target locus by a menin inhibitor and by none of three DOT1L inhibitors - is measured in KMT2A::MLLT3, an AML line with a retained wild-type KMT2A allele and an FLT3-ITD [E2]. This project's recurring finding is that its load-bearing experiments are partner-imported, and this is another one. **2. The RS4;11 arm has no menin inhibitor at all**, so even the mark channel cannot be put on a menin-versus-DOT1L contrast in the index fusion from this series. **3. The two arms are on different genome builds.** `Genome_build: GRCh37/hg19 (H3K79me2 ChIPseq); GRCh38/hg38 (DOT1L & Menin ChIPseq)` [E1]. Even within MOLM13, joining the factor occupancy to the mark requires a liftover, which nothing in this beat does and which no entry in this corpus should do silently. **What DOES exist for the index fusion in this series**, and it is worth having: a dm6-spike-in-normalised H3K79me2 time course in RS4;11 under EPZ5676 and compound 10, already curated as `ana-the-spike-in-ruler-is-deposited-in-plain-text-and-the-global-change-is-5-to-23-fold`.

Go deeper
Absence class

Never-measured, not measured-and-unreported. A MEN1 ChIP in RS4;11 under VTP-50469 would be a separate immunoprecipitation from a separate culture; nothing in the deposit or the paper suggests one was performed and withheld. The nearest existing measurement is GSE127507/GSE127508 (Krivtsov et al. 2019), which DOES have RS4;11 under VTP-50469 - and this corpus already records that it lacks a matched VTP input for RS4;11, in absence-q118-rs4-11-and-ml-2-have-no-vtp-50469-input-in-gse127507. So there are now two separate deposits, from overlapping author groups, each of which would answer the index-fusion question but for one missing sample type, and the missing sample types are different.

A brief is a starting point, not a finding

The premise was checkable in one fetch of the series record. It is recorded here rather than quietly worked around, because a beat that scopes silently around a wrong premise leaves the premise in circulation. The standing instruction is that when a brief conflicts with the literature, the literature wins and the beat says so.

Facts and figures
absence classnever-measured
what is absentMEN1 ChIP and DOT1L ChIP in RS4;11 (KMT2A::AFF1), and any VTP-50469 arm in RS4;11, within GSE149183
what would close itOne MOLM13-sized experiment in RS4;11: MEN1 ChIP under DMSO and VTP-50469 and at least one DOT1L inhibitor, ideally with the dm6 spike-in the same lab already runs. Five to six libraries.
geoGSE149183
correctsthe Q136 dispatch brief, which stated the MEN1/DOT1L head-to-head was available in RS4;11 as well as MOLM13
shown for fusionKMT2A::MLLT3 in MOLM13 for the factor ChIPs; KMT2A::AFF1 in RS4;11 for H3K79me2 only, and with no menin-inhibitor arm
sample map
{
  "MOLM13_factor_chip": [
    "GSM4491616",
    "GSM4491617",
    "GSM4491618",
    "GSM4491619",
    "GSM4491620",
    "GSM4491621",
    "GSM4491622",
    "GSM4491623",
    "GSM4491624",
    "GSM4491625"
  ],
  "MOLM13_h3k79me2": [
    "GSM4491626",
    "GSM4491627",
    "GSM4491628",
    "GSM4491629",
    "GSM4491630",
    "GSM4491631",
    "GSM4491632",
    "GSM4491633",
    "GSM4491634"
  ],
  "RS4_11_h3k79me2": [
    "GSM4491635",
    "GSM4491636",
    "GSM4491637",
    "GSM4491638",
    "GSM4491639",
    "GSM4491640",
    "GSM4491641",
    "GSM4491642",
    "GSM4491643"
  ],
  "inputs": [
    "GSM4491615",
    "GSM4491644",
    "GSM4491645"
  ]
}
genome builds
{
  "MEN1_and_DOT1L_ChIP": "GRCh38/hg38",
  "H3K79me2_ChIP": "GRCh37/hg19"
}
How this could bite us back
  • Every pharmacodynamic conclusion in this beat is measured in an AML line with a retained wild-type KMT2A allele. Infant KMT2A::AFF1 ALL - the disease this project exists for - differs in lineage, in methylome and in partner, and nothing here has been shown in it.
Still open
  1. Two deposits each nearly answer the index-fusion question and each lacks a different single sample type - GSE127507 lacks the RS4;11 VTP-50469 input, GSE149183 lacks the RS4;11 MEN1 ChIP entirely. Could the two be combined, or does the build and protocol difference make that a fiction?
  2. Does GSE149185, the SuperSeries, contain further SubSeries with an RS4;11 menin-inhibitor arm? It was not enumerated here and that is one fetch.
Every source behind this entry (2)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset

    All 31 GSE149183 sample records were fetched: MEN1 and DOT1L ChIP exist only for MOLM13 (GSM4491616-GSM4491625); RS4;11 contributes nine H3K79me2 samples (GSM4491635-GSM4491643) under DMSO, EPZ5676 and compound 10 at d3/d6/d9 plus one input, and no VTP-50469 sample of any kind. The series overall_design and the per-sample Genome_build lines state both facts directly.

    What this evidence cannot show

    Enumerated from the GEO `targ=self&form=text&view=brief` record for each of the 31 accessions individually, not from the series page's sample table. The SRA run set (SRP258055) was not enumerated, so an undeposited RS4;11 VTP library cannot be formally excluded - but no sample record references one.

    Evidence that argues against this (1)
    • GSE149183 is a SubSeries of GSE149185. The SuperSeries may contain further SubSeries (for example RNA-seq or in vivo arms) that were not enumerated here, so this absence is bounded to GSE149183 and not to the whole study. NCBI Gene Expression Omnibus 2020
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    MOLM-13 is KMT2A::MLLT3 with a heterozygous FLT3-ITD and a retained wild-type KMT2A allele; RS4;11 is KMT2A::AFF1, in frame, B-lineage.

    What this evidence cannot show

    Cellosaurus records the fusion but not this deposit's culture conditions; the lines were not re-authenticated by the depositors in the GEO record.

    Evidence that argues against this (1)
    • None; both fusions are supported by multiple independent PubMed references in their Cellosaurus entries. Cellosaurus (SIB) 2026
Connected to